Survival Analysis Report

Executive Summary

This report aims to demonstrate the application of a Cox Proportional Hazard regression model to examine how four key covariates — Sex, Age at Diagnosis, Disease Duration at Diagnosis, and ACE-III Score — affect the survival probability of individuals across various dementia diagnoses. Additionally, a log normal regression was implemented to identify and order the level of impact from the same variables in the cox regression as well as more additional ones like cognitive and behavioral has on survival, both holistically across all dementia types and within each specific dementia diagnosis. It was found that the holistic approach has labelled the disease duration at diagnosis and some sub categories under ACE-III and Cambridge Behavioural Inventory (CBI) scores as significant factors, but under each different diagnosis types, the factors would vary in importance.

Introduction

Research on dementia survival has led researchers, like the client, to note various factors that potentially impact survivability. Understanding these factors is crucial for improving patient care and tailoring interventions to individual needs. This report focuses on two statistical approaches to analyze survival probabilities in dementia patients: the Cox Proportional Hazards model and the log-normal regression model. The Cox model, a widely used method in survival analysis, examines the relationship between survival time and key covariates, providing insights into how these factors affect patient outcomes. The log-normal regression model, which includes the same covariates as the Cox model along with additional variables, offers a comprehensive view of the predictors influencing survival.

Client’s Aims

The first objective of this report is to produce an interactive survival graph that provides a comprehensive exploration of the time from formal diagnosis to death based on four key covariates: age at diagnosis, sex, disease duration at diagnosis, and overall total cognitive performance measured by the ACE-III test. These covariates were selected by the client based on their expertise and the availability of data. The second aim is to explore additional variables, including those from the CBI carer questionnaire, family history of dementia as defined by a Goldman Score, and cognitive abilities. The client suspects these variables may influence survival rates and wants more clarity on the extent of the impact each factor has on survival.

Interactive Survival Graph

The interactive survival graph is designed to facilitate the analysis of different patient groups by inputting their specific data. This feature allows users to observe and compare how various factors impact survival rates across different groups. As shown in the interactive graph below, users can select patient attributes such as age at diagnosis, disease duration, ACE-III total scores, and gender, along with the specific dementia diagnoses. By adjusting these parameters, the graph dynamically illustrates the survival probabilities over time, enabling a detailed comparison between two distinct patient groups.

Model

To meet the client’s objective of understanding how various covariates affect different dementia patient groups, we implemented a Cox Proportional Hazards regression model for each type of dementia, serving as the foundation of the interactive survival graph. This approach was informed by previous client analyses that highlighted statistically significant survival differences across dementia diagnoses. Each model accounted for age at diagnosis, sex, disease duration at diagnosis, and overall cognitive performance as assessed by the ACE-III test. While the Cox regression model does not assume any underlying data distribution, it does require other assumptions to be satisfied (details in the appendix).

Interactive Graph for Comparasion

Patient Group 1
Patient Group 2

Significance of Factors’ Impact

Given the absence of censored data, a parametric log-normal regression model was applied under two distinct scenarios: one incorporating diagnosis types as a variable to assess overall impact, and another providing a more detailed analysis of each individual diagnosis.

The dataset demonstrated multicollinearity, particularly evident with “ACE-III Total” which aggregates all other ACE-III component scores. Such multicollinearity can obscure reliable coefficient estimation and complicate interpretation. To mitigate this, the Variance Inflation Factor (VIF) was employed to identify high multicollinearity, resulting in the exclusion of “ACE-III Total” from the predictors (See Appendix for more details).

For effective comparison of the different covariates importance, scaling of the data was essential to standardize the numerical predictors. Although this scaling complicates the interpretation of individual factors’ direct impact, it facilitates an assessment of the effect size attributed to each variable based on the magnitude of its coefficient. Variables with a greater coefficient magnitude exert a more pronounced influence on the time from diagnosis to death.

Significance of factors across all dementia diagnosis

A comprehensive test was performed to evaluate the overall significance of observations across each diagnostic category, analyzing the effects of outcomes in relation to the 18 factors provided by the client. Furthermore, the model incorporated diagnosis type as a covariate to enhance the understanding of its influence on the outcomes.

In the comprehensive model that accounted for all variables, a bar plot was used to visualize the significant predictors. These included: “Disease duration at diagnosis,” indicating the number of years a person has experienced symptoms prior to their clinic visit; “ACEIII::SubtotalFluency,” which assesses verbal fluency integrity using a cognitive screening tool; “ACEIII::SubtotalVisuospatial,” evaluating visuospatial abilities; and “CBI::PercentSelfCareFCorrected,” reflecting the percentage of the maximum achievable score in self-care as assessed by The Cambridge Behavioural Inventory-Revised (CBI-R), with higher scores indicating superior self-care performance.

The type of effect each variables has on the outcome is depicted in the barchart; an orange bar signifies a positive impact on the model, while those in blue indicate a negative impact.

It was also observed that when different diagnosis types were incorporated into the model, the FTD-MND type emerged as a significant factor, suggesting that having FTD-MND has a statistically measurable impact on the outcome compared to the base level, which is the AD type. Given that FTD-MND was not the primary focus of the investigation, it was excluded from the chart. Full details of this decision are documented in the appendix.

Significance of factors by dementia diagnosis

While the initial model provided a comprehensive overview of general survival impacts, an alternative approach was to evaluate the effects of each factor within specific diagnosis types. Consequently, individual models were fitted for each type of dementia diagnosis. However, certain diagnosis groups—specifically nfvPPA + Parkinson’s plus, PSP, and svPPA—suffered from insufficient observations, which likely led to over-fitting. Over-fitting tend to manifested as models capturing noise rather than true underlying patterns, exacerbated by the fact that the 18 covariates used in each model sometimes exceeded the number of observations available for these diagnosis types (refer to Appendix for details). As a result, the analysis was concentrated on dementia groups with adequate data, identifying significant covariates that influenced the time from diagnosis to death.

Significant variables impacting the survival times for each dementia group were identified through visualization in the analysis. Key findings include:

  • Alzheimer’s Disease (AD): “ACEIII::SubtotalVisuospatial” positively affects survival, whereas “EducationYearsTotal” and “Age at diagnosis” have a detrimental impact.

  • Behavioral Variant Frontotemporal Dementia (bvFTD): Positive influences are noted from “CBI::PercentMemory”, “ACEIII::SubtotalVisuospatial”, and “ACEIII::SubtotalFluency”; “CBI::PercentSelfCare” exerts a negative effect.

  • Corticobasal Syndrome (CBS): “ACEIII::SubtotalAttention” enhances survival prospects, while “CBI::PercentEating” and “CBI::PercentAbnormal” reduce them.

  • FTD-Motor Neuron Disease (FTD-MND): Positive contributions come from “Sex (Male 1, Female 2)” and “Disease duration at diagnosis”; “CBI::PercentAbnormal” negatively influences outcomes.

  • Logopenic Variant Primary Progressive Aphasia (lvPPA): “CBI::PercentSelfCare” and “ACEIII::SubtotalLanguage” are beneficial; “CBI::PercentAbnormal” is harmful.

  • Nonfluent Variant Primary Progressive Aphasia (nfvPPA): Positives include “Sex (Male 1, Female 2)”, “CBI::PercentEating”, and “ACEIII::SubtotalFluency”; “ACEIII::SubtotalAttention” impacts negatively.

The recurrent significance of variables such as “Age at diagnosis”, and “Disease duration at diangosis”, several ACEIII scores and CBI scores across most dementia groups underscores their importance to time from diagnosis to death, as well as supporting the client’s choice of covaraites in the interactive survival graph.

Conclusion

In conclusion, this report has successfully met the client’s two primary objectives. Firstly, it offers an interactive survival chart grounded on the Cox Proportional Hazard model, illustrating the interplay among the four key covariates. Secondly, two distinct approaches were implemented to examine the impact of the 18 covariates of interest to the client. While the first approach provides a comprehensive overview of the current data, the second approach is designed to be concise and can be easily scalable through the inclusion additional diagnosis types as more data becomes available.

Appendix

Testing Proportional Harzard

Covariate Chisq Degrees of Freedom p-value
Age at diagnosis 1.24 1 0.266
Sex 1.35 1 0.246
Disease duration at diagnosis 0.17 1 0.682
ACE-III Total 1.09 1 0.297
GLOBAL 3.25 4 0.517
Covariate Chisq Degrees of Freedom p-value
Age at diagnosis 0.42 1 0.517
Sex 6.61 1 0.010
Disease duration at diagnosis 0.41 1 0.523
ACE-III Total 1.02 1 0.312
GLOBAL 9.12 4 0.058
Covariate Chisq Degrees of Freedom p-value
Age at diagnosis 0.05 1 0.826
Sex 1.88 1 0.170
Disease duration at diagnosis 0.01 1 0.933
ACE-III Total 1.64 1 0.200
GLOBAL 3.87 4 0.424
Covariate Chisq Degrees of Freedom p-value
Age at diagnosis 0.36 1 0.548
Sex 0.05 1 0.824
Disease duration at diagnosis 0.84 1 0.358
ACE-III Total 4.50 1 0.034
GLOBAL 6.95 4 0.139
Covariate Chisq Degrees of Freedom p-value
Age at diagnosis 0.00 1 0.992
Sex 0.12 1 0.731
Disease duration at diagnosis 2.92 1 0.088
ACE-III Total 0.01 1 0.905
GLOBAL 3.86 4 0.425
Covariate Chisq Degrees of Freedom p-value
Age at diagnosis 0.08 1 0.782
Sex 1.50 1 0.220
Disease duration at diagnosis 4.85 1 0.028
ACE-III Total 1.57 1 0.211
GLOBAL 10.68 4 0.030
Covariate Chisq Degrees of Freedom p-value
Age at diagnosis 0.12 1 0.734
Sex 0.05 1 0.832
Disease duration at diagnosis 1.25 1 0.263
ACE-III Total 0.78 1 0.377
GLOBAL 2.34 4 0.674
Covariate Chisq Degrees of Freedom p-value
Age at diagnosis 0.15 1 0.702
Sex 0.07 1 0.796
Disease duration at diagnosis 0.14 1 0.705
ACE-III Total 2.46 1 0.117
GLOBAL 2.87 4 0.579
Covariate Chisq Degrees of Freedom p-value
Age at diagnosis 0.12 1 0.734
Sex 0.00 1 0.972
Disease duration at diagnosis 0.39 1 0.530
ACE-III Total 0.76 1 0.384
GLOBAL 3.47 4 0.483

Checking multi-collinearity

Number of observations per dementia diagosis type

Table 10 : Distribution of Diagnosis Types
Diagnosis Type Number of Observations
AD 49
bvFTD 61
CBS 25
FTD-MND 35
lvPPA 27
nfvPPA 29
nfvPPA + Parkinson’s plus 13
PSP 21
svPPA 21

Results from all dementia diagnosis

Table 11 : Distribution of Diagnosis Types
Coefficient Std. Error z value p value
(Intercept) 1.6287042 0.0991719 16.4230431 0.0000000
Diagnosis_nameFTD-MND -1.3149343 0.1465575 -8.9721390 0.0000000
Log(scale) -0.6053870 0.0421825 -14.3516303 0.0000000
Diagnosis_namenfvPPA + Parkinson’s plus -0.2807378 0.1883152 -1.4907868 0.1360175
Diagnosis_namelvPPA -0.2083461 0.1400319 -1.4878470 0.1367912
Diagnosis_namenfvPPA -0.1942113 0.1576234 -1.2321225 0.2179033
ACEIII::SubtotalFluency 0.1758444 0.0477791 3.6803610 0.0002329
ACEIII::SubtotalVisuospatial 0.1300042 0.0559673 2.3228600 0.0201867
CBI::PercentSelfCareFCorrected -0.1291601 0.0496838 -2.5996408 0.0093321
Sex (Male 1, Female 2) 2 0.1117217 0.0726728 1.5373244 0.1242139
CBI::PercentMemoryFCorrected 0.1082240 0.0557348 1.9417675 0.0521652
ACEIII::SubtotalMemory -0.0980801 0.0644657 -1.5214318 0.1281515
Diagnosis_namebvFTD -0.0789263 0.1392427 -0.5668258 0.5708326
Disease duration at diagnosis 0.0776235 0.0361458 2.1475100 0.0317527
CBI::PercentMotivationFCorrected -0.0761239 0.0533105 -1.4279338 0.1533109
Age at diagnosis -0.0621485 0.0376262 -1.6517348 0.0985886
CBI::PercentAbnormalFCorrected -0.0554625 0.0575207 -0.9642179 0.3349367
Diagnosis_namesvPPA 0.0506025 0.1866061 0.2711727 0.7862582
Diagnosis_namePSP -0.0503900 0.1746879 -0.2884574 0.7729967
ACEIII::SubtotalAttention 0.0499284 0.0620538 0.8045995 0.4210508
ClinicalAssessment::GoldmanScore 0.0471112 0.0355100 1.3267019 0.1846073
CBI::PercentSleepFCorrected -0.0468760 0.0396049 -1.1835893 0.2365757
Diagnosis_nameCBS -0.0467270 0.1698713 -0.2750732 0.7832600
CBI::PercentEverydayFCorrected -0.0387882 0.0580971 -0.6676443 0.5043607
CBI::PercentMoodFCorrected 0.0335674 0.0504296 0.6656296 0.5056479
CBI::PercentBeliefsFCorrected 0.0168583 0.0374950 0.4496140 0.6529888
EducationYearsTotal 0.0100432 0.0371356 0.2704470 0.7868164
CBI::PercentStereotypicalFCorrected 0.0074927 0.0546087 0.1372075 0.8908668
ACEIII::SubtotalLanguage -0.0072250 0.0622213 -0.1161174 0.9075595
CBI::PercentEatingFCorrected -0.0030126 0.0564909 -0.0533295 0.9574693

Results from each dementia diagnosis

Coefficient Std. Error z value p value
(Intercept) 1.9750823 0.1304882 15.1361021 0.0000000
Log(scale) -1.0695636 0.1010153 -10.5881398 0.0000000
ACEIII::SubtotalVisuospatial 0.3234895 0.0731973 4.4194162 0.0000099
CBI::PercentEatingFCorrected 0.2350261 0.1710496 1.3740229 0.1694346
Age at diagnosis -0.2089849 0.0555404 -3.7627564 0.0001681
CBI::PercentSelfCareFCorrected -0.1704212 0.1295625 -1.3153584 0.1883895
EducationYearsTotal -0.1431628 0.0674650 -2.1220319 0.0338351
Sex (Male 1, Female 2) 2 0.1286976 0.1178721 1.0918412 0.2749029
ACEIII::SubtotalLanguage -0.1268934 0.1546210 -0.8206738 0.4118321
Disease duration at diagnosis -0.1220367 0.0729627 -1.6725902 0.0944080
ACEIII::SubtotalAttention 0.1068610 0.1215558 0.8791107 0.3793413
CBI::PercentAbnormalFCorrected -0.0803797 0.1685262 -0.4769568 0.6333929
ACEIII::SubtotalFluency -0.0736457 0.0967825 -0.7609406 0.4466925
ClinicalAssessment::GoldmanScore 0.0730915 0.1044951 0.6994732 0.4842564
CBI::PercentMotivationFCorrected -0.0712989 0.1096646 -0.6501540 0.5155927
CBI::PercentSleepFCorrected -0.0653759 0.0751421 -0.8700306 0.3842837
CBI::PercentStereotypicalFCorrected -0.0629505 0.1336806 -0.4709021 0.6377107
ACEIII::SubtotalMemory 0.0517398 0.1237023 0.4182605 0.6757567
CBI::PercentMemoryFCorrected -0.0411183 0.1035539 -0.3970718 0.6913145
CBI::PercentMoodFCorrected 0.0392490 0.0953593 0.4115904 0.6806397
CBI::PercentBeliefsFCorrected -0.0282680 0.0972206 -0.2907617 0.7712336
CBI::PercentEverydayFCorrected 0.0050338 0.1057047 0.0476213 0.9620180
Coefficient Std. Error z value p value
(Intercept) 1.5303344 0.1170029 13.0794614 0.0000000
Log(scale) -0.8642163 0.0905357 -9.5455812 0.0000000
ACEIII::SubtotalVisuospatial 0.3778851 0.1298433 2.9103170 0.0036106
Sex (Male 1, Female 2) 2 0.2886560 0.1653018 1.7462366 0.0807698
ACEIII::SubtotalFluency 0.2565609 0.0770503 3.3297847 0.0008691
CBI::PercentMemoryFCorrected 0.2381783 0.0976174 2.4399175 0.0146906
CBI::PercentMotivationFCorrected -0.1902755 0.1033317 -1.8414041 0.0655624
CBI::PercentSelfCareFCorrected -0.1681407 0.0732513 -2.2953961 0.0217104
CBI::PercentEverydayFCorrected 0.1342125 0.1096458 1.2240559 0.2209312
CBI::PercentMoodFCorrected 0.1280998 0.0833549 1.5367993 0.1243425
CBI::PercentAbnormalFCorrected -0.1038375 0.0903971 -1.1486811 0.2506875
Age at diagnosis -0.0719060 0.0708841 -1.0144167 0.3103840
Disease duration at diagnosis 0.0638477 0.0551819 1.1570405 0.2472558
ACEIII::SubtotalMemory 0.0593544 0.1088862 0.5451044 0.5856817
ACEIII::SubtotalLanguage -0.0566174 0.1236980 -0.4577070 0.6471629
CBI::PercentStereotypicalFCorrected -0.0359267 0.0909457 -0.3950342 0.6928176
CBI::PercentBeliefsFCorrected -0.0344206 0.0472027 -0.7292072 0.4658749
EducationYearsTotal -0.0305323 0.0749984 -0.4071060 0.6839301
ClinicalAssessment::GoldmanScore 0.0272971 0.0433647 0.6294791 0.5290355
CBI::PercentSleepFCorrected -0.0235812 0.0687799 -0.3428496 0.7317116
ACEIII::SubtotalAttention -0.0116516 0.1224109 -0.0951844 0.9241683
CBI::PercentEatingFCorrected -0.0043376 0.0854637 -0.0507537 0.9595217
Coefficient Std. Error z value p value
Log(scale) -2.5074648 0.1414214 -17.7304536 0.0000000
(Intercept) 1.2601504 0.0813703 15.4866045 0.0000000
ACEIII::SubtotalAttention 0.6429914 0.0924133 6.9577794 0.0000000
CBI::PercentAbnormalFCorrected -0.6081172 0.1485806 -4.0928443 0.0000426
CBI::PercentEatingFCorrected -0.5693867 0.2192432 -2.5970548 0.0094027
Age at diagnosis -0.3749717 0.0397285 -9.4383467 0.0000000
ACEIII::SubtotalFluency -0.3596098 0.0783762 -4.5882543 0.0000045
ACEIII::SubtotalLanguage 0.2561218 0.0901964 2.8396030 0.0045170
Disease duration at diagnosis 0.2549515 0.0475947 5.3567202 0.0000001
ACEIII::SubtotalMemory -0.2291749 0.1057606 -2.1669206 0.0302409
ACEIII::SubtotalVisuospatial -0.1924689 0.0468063 -4.1120264 0.0000392
CBI::PercentMotivationFCorrected 0.1909696 0.0645523 2.9583709 0.0030927
CBI::PercentMemoryFCorrected 0.1740816 0.0795944 2.1871083 0.0287346
CBI::PercentStereotypicalFCorrected 0.1740530 0.1266023 1.3748014 0.1691930
CBI::PercentBeliefsFCorrected 0.1633968 0.0623610 2.6201742 0.0087885
CBI::PercentSleepFCorrected 0.1619915 0.0562377 2.8804782 0.0039707
ClinicalAssessment::GoldmanScore 0.1399078 0.0498079 2.8089447 0.0049704
CBI::PercentEverydayFCorrected -0.1254057 0.0928637 -1.3504274 0.1768789
Sex (Male 1, Female 2) 2 -0.0498895 0.0914201 -0.5457171 0.5852604
CBI::PercentSelfCareFCorrected 0.0191061 0.0787067 0.2427508 0.8081984
EducationYearsTotal 0.0189151 0.0368374 0.5134742 0.6076196
CBI::PercentMoodFCorrected 0.0076438 0.0579347 0.1319384 0.8950330
Coefficient Std. Error z value p value
Disease duration at diagnosis 1.2358327 0.1325214 9.3255344 0.0000000
Log(scale) -1.1623290 0.1195229 -9.7247417 0.0000000
Sex (Male 1, Female 2) 2 0.9278389 0.2189638 4.2374072 0.0000226
CBI::PercentAbnormalFCorrected -0.7573791 0.1153260 -6.5672913 0.0000000
(Intercept) 0.7440880 0.1048109 7.0993354 0.0000000
ACEIII::SubtotalFluency 0.5266696 0.0878011 5.9984373 0.0000000
CBI::PercentSelfCareFCorrected -0.4768137 0.1604920 -2.9709504 0.0029688
CBI::PercentEverydayFCorrected 0.4269279 0.1913710 2.2308912 0.0256883
CBI::PercentSleepFCorrected 0.3911099 0.0964967 4.0530894 0.0000505
CBI::PercentMotivationFCorrected 0.3800288 0.1054487 3.6039195 0.0003135
CBI::PercentStereotypicalFCorrected 0.3773127 0.1193792 3.1606229 0.0015743
ACEIII::SubtotalVisuospatial -0.3649682 0.1723536 -2.1175557 0.0342127
CBI::PercentEatingFCorrected -0.2980087 0.1297252 -2.2972311 0.0216056
ClinicalAssessment::GoldmanScore -0.2946542 0.0716841 -4.1104531 0.0000395
ACEIII::SubtotalLanguage 0.2907866 0.1230162 2.3638076 0.0180882
ACEIII::SubtotalMemory 0.2434248 0.1375060 1.7702854 0.0766796
Age at diagnosis -0.2315923 0.1109884 -2.0866352 0.0369211
CBI::PercentMoodFCorrected 0.2129421 0.1105221 1.9266921 0.0540180
CBI::PercentBeliefsFCorrected -0.1845847 0.1034678 -1.7839817 0.0744266
EducationYearsTotal 0.1306507 0.0952751 1.3712994 0.1702816
CBI::PercentMemoryFCorrected -0.0740127 0.1231594 -0.6009502 0.5478731
ACEIII::SubtotalAttention -0.0224537 0.1301940 -0.1724635 0.8630732
Coefficient Std. Error z value p value
(Intercept) 1.8026096 0.1209581 14.9027656 0.0000000
Log(scale) -1.6674571 0.1360828 -12.2532574 0.0000000
CBI::PercentSelfCareFCorrected 1.4020805 0.2238965 6.2621819 0.0000000
CBI::PercentAbnormalFCorrected -0.6431153 0.1610063 -3.9943497 0.0000649
ACEIII::SubtotalLanguage 0.6034872 0.0958470 6.2963588 0.0000000
Age at diagnosis -0.5169174 0.0752091 -6.8730720 0.0000000
ACEIII::SubtotalMemory -0.3792927 0.1358856 -2.7912649 0.0052502
Sex (Male 1, Female 2) 2 0.3643523 0.1147199 3.1760158 0.0014931
Disease duration at diagnosis 0.3573427 0.1005937 3.5523380 0.0003818
CBI::PercentStereotypicalFCorrected 0.3484087 0.1180543 2.9512583 0.0031648
CBI::PercentMoodFCorrected -0.3461832 0.1049336 -3.2990700 0.0009701
CBI::PercentEverydayFCorrected -0.3000232 0.0746338 -4.0199351 0.0000582
ACEIII::SubtotalFluency 0.2770906 0.0874911 3.1670741 0.0015398
ClinicalAssessment::GoldmanScore -0.2643619 0.0837359 -3.1570927 0.0015935
CBI::PercentMemoryFCorrected -0.2117017 0.1189139 -1.7802942 0.0750278
CBI::PercentBeliefsFCorrected -0.1757078 0.1523796 -1.1530929 0.2488722
CBI::PercentSleepFCorrected -0.1253179 0.0928932 -1.3490534 0.1773198
ACEIII::SubtotalAttention 0.0785194 0.1326390 0.5919781 0.5538653
CBI::PercentMotivationFCorrected 0.0647494 0.1056918 0.6126246 0.5401246
EducationYearsTotal 0.0539759 0.0587912 0.9180949 0.3585692
ACEIII::SubtotalVisuospatial -0.0466682 0.1493251 -0.3125279 0.7546394
CBI::PercentEatingFCorrected 0.0387899 0.1308310 0.2964889 0.7668568
Coefficient Std. Error z value p value
(Intercept) 1.3547831 0.3874500 3.4966660 0.0004711
Log(scale) -1.0773595 0.1313064 -8.2049254 0.0000000
CBI::PercentEatingFCorrected 1.0274441 0.4081958 2.5170376 0.0118346
ACEIII::SubtotalAttention -0.7660366 0.3165456 -2.4199878 0.0155210
CBI::PercentSelfCareFCorrected -0.5565359 0.6558621 -0.8485563 0.3961282
ACEIII::SubtotalFluency 0.5051080 0.1546901 3.2652894 0.0010935
CBI::PercentAbnormalFCorrected -0.4995772 0.2653886 -1.8824364 0.0597768
Sex (Male 1, Female 2) 2 0.4792632 0.1926135 2.4882116 0.0128387
Disease duration at diagnosis -0.2879376 0.1480755 -1.9445316 0.0518314
ACEIII::SubtotalVisuospatial 0.2876360 0.2855660 1.0072491 0.3138151
CBI::PercentMoodFCorrected 0.2815494 0.1576189 1.7862668 0.0740561
CBI::PercentStereotypicalFCorrected -0.2402609 0.2051067 -1.1713945 0.2414402
CBI::PercentMotivationFCorrected -0.2310093 0.1794759 -1.2871324 0.1980481
ACEIII::SubtotalMemory 0.1961746 0.2373103 0.8266587 0.4084305
CBI::PercentBeliefsFCorrected 0.1676204 0.9042686 0.1853658 0.8529421
ACEIII::SubtotalLanguage -0.1363656 0.2832484 -0.4814346 0.6302077
ClinicalAssessment::GoldmanScore 0.1231585 0.2982228 0.4129749 0.6796250
CBI::PercentSleepFCorrected 0.0974343 0.1751731 0.5562172 0.5780624
Age at diagnosis -0.0701088 0.1005063 -0.6975563 0.4854547
CBI::PercentMemoryFCorrected 0.0644035 0.1824586 0.3529758 0.7241066
EducationYearsTotal -0.0228537 0.1797444 -0.1271453 0.8988254
CBI::PercentEverydayFCorrected -0.0154270 0.2513615 -0.0613739 0.9510614